Security audit
Security checks for vulnerabilities and agentic risk
This is a documentation helper for OpenAPI and Swagger work, with some overly broad activation wording but no hidden execution, persistence, credential use, or data exfiltration behavior.
This skill appears safe to install for OpenAPI and Swagger documentation tasks. Be aware that its trigger wording is broad, so you may want to invoke it explicitly only when you are working on API documentation or validation, and avoid relying on automatic activation for general backend questions.
The trigger sentences are broad, repetitive, and not tightly scoped to explicit user intent, which can cause the skill to activate in situations where the user did not clearly request it. In an agent setting, ambiguous activation increases the chance of misrouting requests, applying the wrong workflow, or exposing unrelated context to the skill.
The trigger phrases are very broad and generic, which can cause the skill to activate outside the user's intended scope. In an agent environment, ambiguous activation increases the chance of misrouting requests, unnecessary tool use, or applying documentation-generation behavior to unrelated prompts, which can degrade safety and reliability.
The description says to use the skill when a user asks for broad topics like 'software-and-data' or merely 'needs a practical workflow, artifact, checklist, analysis, or implementation support,' which overlaps with many ordinary technical requests. It does not clearly distinguish when this skill should activate versus other software-related skills, increasing the risk of unintended invocation.
The trigger keywords include very broad terms such as 'software-and-data' and generic API-related phrases, which can cause the skill to activate for loosely related requests. Overbroad routing can expose users to irrelevant or misleading workflows, increase prompt-injection surface from mismatched contexts, and interfere with safer or more appropriate skills.
The skill description uses broad activation criteria such as general software/API documentation topics, which can cause the skill to trigger during ordinary technical conversations not specifically requesting this workflow. Over-broad routing increases the chance of unintended instruction injection into unrelated sessions and can override more appropriate skills or baseline behavior.
The keyword trigger list includes generic terms like 'software-and-data' and broadly used API-related phrases that overlap heavily with normal engineering discussions. This weak specificity can misroute many unrelated prompts into the skill, expanding the attack surface for prompt interference and causing irrelevant or unsafe automated behavior.
The default prompt contains a very broad natural-language trigger phrase tied to common API/documentation terms, which increases the chance the skill is invoked in situations the user did not explicitly intend. In a documentation-focused skill, accidental routing can expose user prompts or cause the agent to apply this skill in unrelated software conversations, creating prompt-scope confusion and increasing the attack surface for downstream prompt injection or mis-execution.
Enabling implicit invocation without clear activation boundaries allows the skill to activate automatically based on ambiguous user language rather than an explicit request. Because this skill targets broad developer workflows, that behavior makes it easier for unrelated conversations to be routed through the skill, increasing the risk of unintended tool use, data exposure to the skill context, and exploitation through crafted prompts that trigger the skill indirectly.
The trigger sentence is so broad and awkwardly templated that it can match routine user phrasing unrelated to explicit invocation, causing the skill to activate in contexts the user did not intend. In a routing or orchestration system, this increases the chance of unsolicited prompt injection exposure, incorrect tool selection, or misleading documentation-oriented responses being inserted into unrelated conversations.
Referencing multiple language-specific skill files without stating that language selection is based on user preference or explicit locale opt-in can lead to unintended language routing. This is primarily a safety and UX control issue: the wrong-language instructions could be surfaced or selected without clear user consent, causing confusion or incorrect execution.
This README is specifically presented in Simplified Chinese and references separate Chinese and English documentation files, but the file itself does not state that language selection is optional or user-driven. That can be interpreted as a locale-specific presentation without explicit opt-in or justification within this file.
The example trigger phrases are phrased as generic help requests and practical workflow asks, which are common in benign day-to-day conversations. If examples are used for activation tuning or matching guidance, they can encourage accidental invocation in contexts where the skill is not the best fit.
The file mixes English skill content with Chinese-language titles and sources, but more importantly the trigger and usage material do not document any language choice or locale behavior for users. If the skill is intended to operate in a specific language or locale context, that constraint is not explicitly offered as a user choice or justified.
